1722
J. Org. Chem. 1982,47, 1722-1724
Preparation of Some Unsaturated Side-Chain Derivatives of Cholesterol Henry W. Kircher* and Fumiko U. Rosenstein Department of Nutrition and Food Science, College of Agriculture, The University of Arizona, Tucson, Arizona 85721 Received S e p t e m b e r 21, 1981
Cholesta-5,22(E),25-trien-3@-01(61, cholesta-5,25-dien-3@-01(7), 5a-cholest-25-en-3@-01(8), cholesta-5,22(E),24-trien-3@-01(9),and cholesta-5,22(E)-dien-3@-01were prepared from the product obtained by reaction of 3/3-acetoxychola-5,23-dien-22-ol with an N,O-ketene acetal.
Cholesterol derivatives with unsaturated side chains are commonly prepared with organometallic or Wittig reagents and suitable C19-CMcarbonyl derivatives.' More recently, allylic rearrangemer~ts~?~ and the ene reaction4y5were used to generate specific chirality at CZO. A Claisen rearrangement was extensively used by Sucrow and m-workers to prepare (24R)- and (24S)-24-ethyl sterols with A25 unsaturation in the side chain.h4 In this paper we describe an application of this rearrangement to the synthesis of a number of cholesterol derivatives. Our starting material was (20S,22R,S)-3P-acetoxychola-5,23-dien-22-01(2)prepared from (20S)-3&acetoxybisnorcho1-5-en-22-al (1)and vinylmagnesium bromide.' Alcohol 2 reacted readily with 1-methoxy-1-(dimethylamino)-1-propene8 in refluxing benzene to give (22E,25R,S)-N~-dimethyl-3~-acetoxycholesta-5,22-dien26-amide (3, Scheme I). Amide 3 was then stepwise hydrogenated in neutral and acid solutions to the A5 (4) and saturated (5) amides, respectively. The three amides were reduced with LAH in THF to hydroxyamines 3a-5a and these in turn oxidized to the N-oxides 3b-5b with H202 in MeOH. From here our methodology diverged from that of Sucrow et al. Pyrolysis of 3b in Me2S06a4 gave poor yields of the desired sterol (6); the main product of the reactions was amine 3a.9 Pyrolysis in pyridine improved the yield of 6 but still gave appreciable amine. Addition of solid alkali10 to the hot pyridine prior to addition of the N-oxides inhibited amine formation and provided the desired unsaturated sterols in good yield and without epimerization at C3. The alkali of choice for conversion of 3b to 6 was Ba(OH),; pyrolysis of N-oxides 4b to 5,25-cholestadienol (7) and 5b to 25-cholestenol (8) went smoothly and more rapidly with KOH. The use of KOH in the pyrolysis of 3b led to various quantities of a byproduct that was shown to be 5,22(E),24-~holestatrienol(9) and which formed by (1)D.M. Piatak and J. Wicha, Chem. Rev., 78,199 (1978). 102,862(1980). (2)M. Tanabe and K. Hayashi, J. Am. Chem. SOC., (3)M. Koreeda, Y.Tanaka, and A. Schwartz, J.Org. Chem., 45,1172 (1980). (4)W. G.Dauben and T. Brookhart, J. Am. Chem. SOC.,103, 237 (1981). (5) A. D. Batchko, D. E. Berger, M. R. Uskovic, and B. B. Snider, J. Am. Chem. SOC.,103,1293 (1981). (6)(a) W. Sucrow and B. Girgensohn, Chem. Ber., 103,750(1970).(b) W.Sucrow, P. P. Caldeira, and M. Slopianka, ibid., 106,2236 (1973).(c) W.Sucrow and M. Slopianka, ibid., 108,3721(1975).(d) W.Sucrow, M. Slopianka, and P. P. Caldeira, ibid., 108,1101 (1975). (7)H. W.Kircher and F. U. Rosenstein, manuscript in preparation. (8) H. Bredereck, F. Effenberger, and H. P. Beyerlin, Chem. Ber., 97, 3081 (1964). (9)Amiies as byproduets of N-oxide pyrolyses constituted 39% of the products when no j3-H is present (A. C. Cope, T. T. Foster, and P. H. Towle, J. Am. Chem. SOC.,71,3929(1949)]and 10-28% of the products in other cases: A. C. Cope, C. L. Bumgardner, and E. E. Schweitzer,ibid., 79,4729(1957);A. C. Cope, E. Ciganek, and N. A. LeBel, ibid., 81,2799 (1959). (10)In contrast, addition of NaOH to the pyrolysis of methylethyl-npropylamine oxide did not change the yield or proportions of olefins in this reaction. A. C. Cope, N. A. LeBel, H.-H. Lee, and W. R. Moore, J. Am. Chem. SOC.,79,4720 (1957).
alkaline isomerization of the 5,22(E),25-triene. The reaction of 6 with KOH in hot pyridine was then used to prepare 9. 5,22(E),25-Cholestatrienol(6) was previously reported to have been synthesized by isomerization of cholesta4,22(E),25-trien-3-one(prepared by a Wittig synthesis) to the 5,22(E),25-trienone followed by LAH reduction and purification by preparative TLC." Since the melting point that we observed for 6 [119-120 "C (air), 121.5122.5 (in vacuo, corr)] was so different from the one reported" (140 "C), we believe the latter to be an isomer, possibly the 20430 derivative,' of 6. All of our chromatographic and spectral data agree with the assigned structure. The physical and spectroscopic properties that we observed for 5,22(E),24-~holestatrienol(9) and its acetate were in accord with those given by Hutchins et a1.,12 and those of 5,25-cholestadienol (7) and its acetate corresponded well to literature values.13 25-Cholestenol(8)has not been reported before either as a synthetic or naturally occurring sterol. Our initial goal in this study was to prepare trans-22dehydrocholesterol (10) for work with DrosophiZa. We had previously prepared 10 by the conventional Wittig reaction, but separation of this compound from 30% of its AZ2cis isomer as acetates was tedious on a gram scale.14 The ketene acetal route described here appeared useful in that it gives only trans-22-dehydro~terols.'~ However, unlike the facile hydrogenations of the (24R)- and (24S)-24-ethyl A22(E)925side chains over a soluble rhodium catalyst to the corresponding A22(E)derivatives,&Sdabsence of an alkyl group at CZ4in triene 6 makes its AZ2bond much more susceptible to reduction, and it was difficult to get good conversion of 6 to 10 over this catalyst. A small amount of 6 was reduced to 10 over a commercial Ni catalyst, but the reaction had to be stopped before all the 6 had been converted to 10 because cholesterol (11; by reduction of the A22 bond) was already apparent by argentation TLC. Reductions of 7 to 11 and 8 to cholestanol(l2) for structure corroboration and confirmation of the normal configuration at C20were straightforward.
Experimental Section Melting points were taken in air or in vacuo in capillary tubes
with a Thomas-Hoover apparatus and are corrected. IR spectra were obtained with a Perkin-Elmer 398 (2.5% solutions in CS2), UV spectra with a Beckamn DU-8 (EtOH), 'H NMR spectra with a Bruker WM-250(CDCl,), mass spectra with a Varian MAT 311 A (direct inlet, 70 eV), optical rotations with a Rudolph DP-06-01 (11)R. Ikan, A. Markus, and E. D. Bergmann, Zsr. J. Chem., 9,259 (1971). (12)R. F.N. Hutchins, M. J. Thompson, and J. A. Svoboda, Steroids, 15, 113 (1970). (13)J. A. Svoboda and M. J. Thompson, J. Lipid Res., 8,1952(1967). (14)F.U. Rosenstein, R. Caruso, and H. W. Kircher, Lipids, 12,297 (1977). , ~ -
(15)W.Sucrow, B. Shubert, W. Richter, and M. Slopianka, Chem. Ber., 104,3689 (1971).
0022-3263/82/1947-1722$01.25/00 1982 American Chemical Society
J,Org. Chem., Vol. 47,No. 9, 1982
Unsaturated Side-Chain Derivatives of Cholesterol
1723
Scheme I a
0
I
6 lO'*ia"( R1H
polarimeter (C3, CHC13),and acetylations with pyridineAczO at room temperature. Chromatography. GC retention times (in minutes) were determined with a 2 m long X 3 mm id, 5% OV-101 column at 250 OC, and a SrWdetector. Relative retention time (RRT) with respect to 11: 3a, 1.95; 4a, 2.13; 5a, 2.16; 6,0.97; 7, 1.07; 8,1.09; 9, 1.18; 10, 0.97; 12, 1.02. TLC was performed with Merck aluminum-backed silica gel plates (0.25 mm) (Sigel) and with these plates dipped in 10% AgN03 in 80% EtOH, air-dried, and activated at 110 OC for 20 min (Ag+ Sigel) in five solvent systems: (1) Sigel, 6:4 hexaneEtOAc, one development; (2) Sigel, 6:4 MeOH-Et,O, one development; (3) Ag+ Sigel, 991 CHC13-Me2C0, 2.5-h development; (4) Ag+ Sigel, 19:l CHC13-Me2C0, 2-h development; (5) Ag+ Sigel, 5:2 hemebenzene, 4 h development. RRT (with respect to llac; ac = steryl acetates): 6ac, 0.11; 7ac, 0.17; 8ac, 0.32; 9ac, 0.28; lOac, 0.65; 12ac, 1.22. In systems 3-5 the top 1 cm of the plate protruded through a slit in an aluminum foil cover of the TLC tank to allow for continuous development. Sigel(100 mesh)
